ORIGINAL 
Annals of Nuclear Medicine Vol. 6, No. 4, 241-246, 1992 
Regional cerebral glucose metabolism in patients with Parkinson's disease with or without dementia 
Masayuki SASAKI,* Yuichi ICHIYA,* Shinichi HOSOKAWA,** Makoto OTSUKA,* Yasuo KUWABARA,* Toshimitsu FUKUMURA,* Motohiro KATO,*** Ikuo GOTO** and Kouji MASUDA* 
*Department of Radiology, **Department of Neurology, ***Department of Neurophysiology, Faculty of Medicine, Kyushu University 
By means of positron emission tomography, the cerebral glucose metabolism in 5 patients with Parkinson's disease with dementia was compared with that in 9 patients without dementia, and that in 5 normal volunteers. The metabolic rates for glucose were measured by placing one hundred regions of interest. In the demented patients, cerebral glucose metabolism was diffusely decreased compared with that of the non-demented patients and the normal controls. The most significant decrease in glucose metabolism was observed in the angular gyrus (49.7% of the normal controls). The glucose metabolism in the cingulate, pre- and postcentral, occipital and subcortical regions was relatively spared (62.1 to 85.5% of the normal controls). In the patients without dementia, the glucose metabolism in each region was not significantly different from that in the normal controls. These results suggest that diffuse glucose hypometabolism in the cerebral cortex may correlate with that of patients with Parkinson's disease with dementia. 
Key words: positron emission tomography, fluorodeoxyglucose, glucose metabolism, Parkinson's disease, dementia 

INTRODUCTION 
PARKlNSON'S DISEASE is a disease entity characterized by extrapyramidal symptoms including a resting tremor, hypokinesia and rigidity. However, recent reports have revealed that approximately one third of patients with Parkinson's disease will develop dementia.1,2 The clinical features of the dementia of Parkinson's disease are similar to those of progressive supranuclear palsy and Huntington's chorea. Their dementia is termed "subcortical dementia" and is differentiated from the cortical dementia representative of Alzheimer's disease.3 In demented patients with Parkinson's disease, cerebral glucose 
Received March 25, 1992, revision accepted May 18, 1992. For reprints contact : Masayuki Sasaki, M.D., Department of Radiology, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812, JAPAN . 
metabolism measured by positron emission tomography is reported to be diffusely decreased in the cerebral cortex.4,5 In the present study, we examined the cerebral glucose metabolism in demented patients with Parkinson's disease, and calculated the metabolic rate for glucose in each small region by placing one hundred regions of interest. We compared them with those in non-demented patients with Parkinson's disease and with those in normal controls, to determine whether there are specific regions of cerebral glucose hypometabolism. 

MATERIALS AND METHODS 
Patients 
Fourteen patients with Parkinson's disease were examined (Table 1). All patients except one (Case 2) had three classic signs of resting tremor, rigidity and akinesia, without pyramidal signs and with good response to levodopa. The severity of their disease was evaluated and classified in stages I, II, III and IV along with their level of functional disability according to Hoehn and Yahr.6 Five patients were demented (PDD); 9 were not (PD), on the basis of neuropsychological evaluation according to Hasegawa's Dementia Scale.7 X-ray computed tomography (XCT) and/or magnetic resonance imaging (MRI) were performed to exclude organic diseases, especially cerebro-vascular diseases. They were also used for the evaluation of cerebral atrophy. The degree of central sulcal widening, an estimation of cortical atrophy, was graded as follows : none, mild (3 mm or less), moderate (4 to 5 mm) and marked (more than 5 mm). Case 2, who had no tremor, was clinically diagnosed as Parkinson's disease, because of the long duration, gradual progression and the effectiveness of levodopa. Before the PET studies were performed, none of the patients received drugs for at least 72 hours. The mean age of the PDD group was 62 yrs (range 56-65 years); that of the PD group was 50.7 yrs (range 37-67 years). The difference between these two groups in the mean age was statistically significant. The clinical stage of the PDD group was III or IV and that of the PD group was I or III. According to the Hasegawa's Dementia Scale, the scores in the PDD patients were lower than 20.0 points and those in the PD patients were 32.5 points. By XCT and/or MRI, mild cortical atrophy was revealed in Case 2, 3, 4 and 5. Five normal female vounteers were also examined for comparison. They were age-matched with the PDD patients and their mean age was 61.5 yrs (range 51-66 years). None of the volunteers had neurological or neuropsychological abnormalities. The absence of organic disorders in them was confirmed by XCT. 
PET system 
We used the 18F-fiuoro-2-deoxy-D-glucose (18F-FDG) method to measure the regional cerebral metabolic rate for glucose (rCMRGlc). Fluorine-18 was produced in a cyclotron BC1710 (The Japan Steel Works Corp.) by the 20Ne(d,a)18F reaction,, and FDG was synthesized by the method reported by Biada et al.8 PET was performed with a HEAD-TOME III (Shimadzu Corporation, Japan) at a spatial resolution of 8.2 mm full width at half maximum (FWHM).9 A transmission scan with a 68Ge/68Ga ring source was obtained for the correction of attenuation. Five contiguous slices were obtained in the orbitomeatal line (OM) +20 mm, +35 mm, +50 mm, +65 mm and +80 mm planes. Two to eight mCi (74 to 296 MBq) of 18F-FDG were administered intravenously, and arterial blood samples were obtained at predetermined intervals from the time of injection until the end of scanning. The rCMRGlc (mg/min/100 ml) was determined from the emission scan (eight minutes scan from commencing 63 minutes after injection), and blood curve data using the model of Phelps et al.,lo later modified by Brooks.11 Literature value for the lumped constant of 0.42 was employed.10 One hundred regions of interest (ROIs) were placed in 22 regions by manual tracing (Fig. 1). Individual regions were combined into anatomical and functional combinations guided by CT and/or MRI, and an atlas of axial tomographic anatomy. The rCMRGlc values were obtained for cortical areas corresponding to the superior frontal (SF). middle frontal (MF), inferior frontal (IF), cingulate (Ci), precentral (Pre), superior temporal (ST), middle temporal (MT), inferior temporal (IT), postcentral (Post), angular (AG), supramarginal (SM), occipital regions (OC) and cuneus (Cu), for subcortical areas corresponding to the caudate nucleus ,(Ca), putamen (Pu), thalamus (Th), hippocampus ,(H) and white matter (WM). The rCMRGlc values for the midbrain (MB), pons (P), cerebellar hemisphere (Ch) and cerebellar vermis (Cv) were also ,obtained. All ROIs consisted of 11-137 mm2. The regions of interest for the global cerebral hemisphere were placed in the slice level OM +50 mm. The rCMRGlc values for the left and the right regions were averaged and used for the rCMRGlc of each region. Statistical analysis for intergroup comparisons of the rCMRGlc were made by Student's t test. 

RESULTS 
The rCMRGlc for each region in the normal controls, the PD group and the PDD group are presented in Table 2. In the PDD group, the rCMRGlc values were decreased in all regions compared with those in both the PD group and the normal controls. In contrast, the rCMRGlc Values for all regions in the PD group were not significantly different from those in the normal controls. The rCMRGlc in the cerebral cortices in the PDD group were diffusely decreased; their values were 3.8-5.61 mg/min/100 ml. The most significant decrease in the rCMRGlc was observed in the angular gyrus; its value was 3.8 mg/min/1OO ml. There was no significant difference between sides (data not shown). The rCMRGlc in the cingulate gyrus, precentral area, postcentral area, cuneus and occipital lobe were decreased and the decrease was statistically significant, but the decrease was not as great as that in the other cerebral cortices. In the subcortical regions in the PDD group, the decreases in the rCMRGlc in the caudate nucleus, putamen, thalamus and hippocampus were significant, but relatively mild compared with those in the cerebral cortices. The rCMRGlc values for the cerebellar hemisphere, cerebellar vermis, pons and midbrain in the PDD group were 4.41-5.40 mg/min/1OO m/. The decreases among them were not statistically significant. However, the rCMRGlc values for the thalamus and cerebellar hemisphere were significantly lower in the PDD group than those in the PD group. The PET images of the PDD (Fig. 2, lower row) show diffuse hypometabolism in the cerebral cortex and relatively preserved pre- and postcentral, occipital regions, striatum, thalamus and cerebellum, in contrast to the normal rCMRGlc pattern of the PD (Fig. 2, upper row). 

DISCUSSION 
The predominant pathology of Parkinson's disease is the degeneration of dopaminergic neurons within the substantia nigra. In demented patients with Parkinson's disease, three types of cortical pathologic changes have been suggested. The first is a brain stem type of Lewy body disease without any change in the cerebral cortex.12 The second type includes neurofibrillary tangles, neuritic plaques and neuronal cell loss which are similar to those seen in Alzheimer's disease.13 The third type is a diffuse Lewy body disease in which both multiple Lewy body and Alzheimer-like pathologic changes can be observed in the cerebral cortex.14 It is well known that the pattern of cerebral glucose metabolism measured by positron emission tomography is associated with pathologic changes in several disorders, as seen in Alzheimer's disease demonstrating predominant hypometabolism in the temporo-parietal cortex.15 Concerning the rCMRGlc in the PDD, several reports have described decreased glucose metabolism in the cerebral cortices. Kuhl et al. first demonstrated uniformly reduced cerebral glucose metabolism in one demented patient.4 They observed severe parietal hypometabolism similar to that seen in Alzheimer's disease. Okada et al. examined seven demented patients and demonstrated diffuse hypometabolism in the cerebral cortex.16 In these reports, the rCMRGlc in cortical lobes was represented by a small ROI placed in one part of each lobe. However, to determine whether there are specific regions of hypometabolism in dementia, glucose metabolism should be measured in numerous small regions. To analyze the pattern of cerebral glucose metabolism, Peppard et al. placed multiple rectangular ROIs along the cerebral cortex and reported observing similar patterns of cortical hypometabolism in both Alzheimer's disease and Parkinson's disease with dementia.5 We examined the cerebral glucose metabolism of the PDD by placing one hundred ROIs, and by tracing every small region, guided by CT and/or MRI, and using an atlas, although it was difficult to identify each gyrus on the PET images because of the limited resolution of our PET scanner. In the present study the rCMRGlc value in the PD group was not significantly different from that in the normal controls in all regions. The rCMRGlc in the caudate nucleus and the putamen were slightly higher than those in the normal controls. but they were not statistically significant. In the PDD group, our results demonstrated diffuse hypometabolism in the cerebral cortex and this was consistent with previous reports.4,16 As mentioned in materials, the mean age of the PDD group was higher than that of the PD group, and the rCMRGlc may decrease with age. However, as our normal controls were age-matched with the PDD group, the decrease in rCMRGlc in the PDD group can not be explained only by the PDD and the PD group age difference. The most significant decrease in rCMRGlc was observed in the angular gyrus. Decreased rCMRGlc in the angular gyrus in the PDD group was also observed by Peppard et al.5 They also suggested that the cortical dysfunction of the PDD has a similar pattern to that of Alzheimer's disease as determined by Q-component analysis.17 We did not analyze the rCMRGlc pattern statistically, but the PET images suggested a similarity between the PDD and Alzheimer's disease (Fig. 2). The neuropsychological dysfunction of the PDD group is suggested to have a relationship with the hypometabolism of the angular gyrus, one of association neocortices, as in Alzheimer's disease. In other association neocortices, the rCMRGlc values for the frontal and temporal cortices were also decreased. It is suggested that cortical pathologic changes in PDD may be classified into three types. It is possible that the three types of PDD have different glucose metabolism patterns. We could not classify our 5 PDD patients into three pathologic types, because their pathologic features have not been verified. Further examinations should be performed to determine whether the glucose metabolism of these types is different. In the PDD group, the rCMRGlc values for the cingulate gyrus, precentral area, postcentral area, cuneus and occipital lobe were not greatly decreased compared with other cerebral cortices. Peppard et al. also demonstrated decreased rCMRGlc in the left perirolandic area.5 In our study, the rCMRGlc values for the right and left precentral areas in the PDD group were 4.93 and 4.77 mg/min/1OO ml, respectively. In the postcentral areas, they were 5.31 mg/min/1OOml on the right side and 5.14 mg/ min/1OOml on the left side. These values were apparently lower than those in the normal controls, but they were relatively spared (64.3 to 74.2% of normal controls) in comparison with the other cerebral cortices. Furthermore, there were no significant differences between sides in the pre- and postcentral areas. The reason for the differences between our data and those of the previous study is not clear. When the XCT and MRI images between the PDD group and the PD group were compared, the incidence of sulcal widening was found to be higher in the PDD group. The partial volume effect due to the cortical atrophy leads to an underestimation of the cortical glucose metabolism. However, the contribution of cortical atrophy in the PDD group was thought to be not so much compared to other groups because the sulcal widening in the PDD group was less than 3 mm. In the present study, we examined the cerebral glucose metabolism of demented patients with Parkinson's disease and calculated the rCMRGlc by placing one hundred regions of interest. In the demented patients, the cerebral glucose metabolism was diffusely decreased in all regions, especially in the angular gyrus, compared with those in the nondemented patients and the normal volunteers. These results suggest that diffuse glucose hypometabolism in the cerebral cortex may correlate with dementia in patients with Parkinson's disease. 

ACKNOWLEDGMENTS 
We thank Dr. Walter J, Russell for his editorial assistance. This work was supported in part by Grants-in Aid from the Ministry of Health and Welfare. 

REFERENCES 
1 . Brown RG, Marsden CD : How common is dementia in Parkinson's disease. The Lancet 2 : 1262-1265, 1984 
2. Lieberman A. Dziatolowski M, Kupersmith M, et al : Dementia in Parkinson disease. Ann Neurol 6: 355-359, 1979 
3. Albert ML, Feldman RG, Willis AL: The 'subcortical dementia' of progressive supranuclear palsy. J Neurol Neurosurgery Psychiatry 37 : 121-130, 1974 
4. Kuhl DE, Metter EJ, Riege WH: Patterns of local cerebral glucose utilization determined in Parkinson's disease by the [18F]fluorodeoxyglucose method. Ann Neurol 15 : 419-424, 1984 
5. Peppard RF, Martin WRW. Clark CM, et al: Cortical glucose metabolism in Parkinson's and Alzheimer's disease J Neurosci Res 27: 561-568, 1990 
6. Hoehn NM, Yahr MD: Parkinsonism: onset, progression, and mortality. Neurology (NY) 17: 427-422, 1966 
7. Hasegawa K, Inoue K, Moriya K : An investigation of dementia rating scale for the elderly. Clinical Psychiatry 16 : 965-969, 1974 
8. Biada GT, Satyamurthy N. Barrio J : The synthesis of 2-[F-18]-fluoro-2-deoxy-D glucose using glycals : a reexamination. J Nucl Med 25: 1327-1334 1984 
9. Kanno I, Miura S, Yamamoto S, et al: Design and evaluation of the positron emission tomography : HEADTOME III. J Comput Assist Tomogr 9: 931-939, 1985 
10. Phelps ME, Huang SC, Hoffman EJ, et al : Tomographic measurements of local cerebral glucose metabolic rate in humans with [18F]2-fluoro-2-deoxy-D-glucose: Validation of method. Ann Neurol 6: 371-388, 1979 
11. Brooks RA: Alternative formula for glucose utilization using labeled deoxyglucose. J Nucl Med 23: 538-539, 1982 
12. Yoshimura M : Pathological basis for dementia in elderly patients with idiopathic Parkinson's disease. Eur Neurol 28 : s29-s35, 1988 
13. Hakim AM, Mathieson G : Dementia in Parkinson's disease : A neuropathologic study. Neurology 29: 1209-1214, 1979 
14. Kosaka K, Yoshimura M, Ikeda K, et al: Diffuse type of Lewy body disease : Progressive dementia with abundant cortical Lewy bodies and senile changes of varying degrees. A new disease Clin Neuropathol 5 : 185-192, 1984 
15. Rapoport SI: Positron emission tomography in Alzheimer's disease in relation to disease pathogenesis: a critical review. Cerebrovasc Brain Metab Rev 3 : 297-335, 1991 
16. Okada J, Peppard R, Calne B: Comparison Study of Positron Emission Tomography, X-ray CT and MRI in parkinsonism with dementia. Nippon Act Radiol 49: 643-656, 1989 
17. Clark C, Carson R, Kessler R, et. al : Alternative statistical methods for the examination of clinical positron emission tomography/fluorodeoxyglucose data. J Cereb Blood Flow Metab 5: 142-150, 1985